Andreas Madsen

Hi, I'm a high school senior and I want to go into ML. Do you think the requirements for a PhD or research labs will get better within the next 4-6 years, and what can I start doing now to improve my chances?

Uh. 5 years ago deep learning was in its infancy. Who can really say what will happen in 5 more years. I think most agree that there is a ML bubble that will eventually burst, so things will change. Better or worse? I don't know.

I love your blog post thanks for the advices and the time for writing all the process, it is interesting to see how hard can be the process of be an independent research... so far which country are you trying to get into to start the Ph.D?

Thanks. It took about 3 days to write :) I've applied in US, UK, Canada, Japan, South Korea, Hong Kong, Singapore, Australia, Switzerland, Denmark, Germany. I might have forgotten a few.

"There is a gigantic gap between what recruiters require, researchers want, and professors provide." Can you elaborate more?

Many researchers are open, some got in with just an MSc. However, HR hard-filters by requiring a PhD, which the professors/universities only give to those who have already learned to research without supervision. At NeurIPS, I saw that these groups are unaware of that mismatch.

I went through something similar this year, where I've been applying for MSc/PhD positions overseas (I'm from Brazil). I'm a software engineer by day and independent researcher by night. I've written 2 papers, one was rejected by NeurIPS 2019, the other I won a prize at UBC. 1/2

The competition is absurd, especially as an international applicant. I've come to the conclusion that it's basically impossible to be accepted without an internal referral. Congrats on the post, I think it will speak to a lot of people :) 2/2